The US military is buying tools to check if AI-generated code is safe

The US military awarded a contract to develop software that can automatically analyze code created by AI. This software will help verify that AI-generated code is secure and reliable for critical defense systems.

For years, the military has struggled with how to trust software, especially when it comes from outside vendors. Now, as AI starts writing code, the problem of trust gets much harder. This contract means the military is building its own tools to check AI-generated code, rather than relying on developers to self-certify its safety. It shifts the burden of proof for AI code safety.